An Integrated Approach for 3D Facies Modeling of Kangan and Dalan Reservoirs, South Pars Gas Field, Persian Gulf

This study focuses on the facies modeling and reservoir characterization of the Permian-Triassic age Dalan and Kangan formations, defined as the main reservoirs in the giant South Pars Gas Field in the Persian Gulf. Based on the main characteristics on petrographical observations, 12 facies are identified and classified into 4 facies associations representing tidal flat (LFAs 1), lagoon (LFAs 2), shoal (LFAs 3), and open marine (LFAs 4) conditions on a carbonate ramp. In uncored wells, a neural network approach (self-organizing maps) was employed to predict litho-facies and litho-facies associations (LFAs). The method was found satisfactory (87.5%) in litho-facies prediction using GR, DT, NPHI, RHOB, and PEF logs. The predicted LFAs were compared with the core-derived facies and rock types to generate a 2D facies model within the sequence stratigraphy framework for geologic modeling and subsequent reservoir simulation. Finally, geostatistical techniques are employed to prepare a 3D facies distribution and depositional model for the entire field. The stochastic simulation method was applied here to simulate and generate the 3D model of 4 major LFAs were involved in the modeling representing LFAs 1, LFAs 2, LFAs 3, and LFAs 4. Facies modeling of the formations indicates a gentle shallowing from zone K4 to zone K3. The connectivity of LFAs 3 is well observed in zone K4, whereas in zone K3 the connectivity of LFAs 2 is evident. Zone K2 is associated with dominant LFAs 3 and minor LFAs 4. The zone K1 is characterized by the dominance of LFAs 1.